On Oct 1, 2007, at 1:08 PM, John Kristoff wrote:
Since this group can be a bit pedantic about terminology, what Danny is referencing is what I was specifically referring to as "resolvers". I totally made up those four categories for my purposes of describing some of the work we were doing as a simple way to describe what we've seen in responses to probes.
Yeah, I'd hope you'd chime in here :-) Sorry to put you on the spot, but I believe data and work are relevant and quite useful in this context. Furthermore, I think it's a useful exercise to classify resolver types, as implications and any recommendation BCPs would perhaps vary widely across each type.
Can you explain what you mean by the 800k being the larger problem? I presume you're referring to the observation that the majority of resolvers on the net are forwarding to that smaller pool.
Yep, precisely.
I might argue that the bigger problem is not with the 800k pool.
Yeah, actually, I could agree with that. One might presuppose that the larger pool is perhaps the result of more default configuration and implementation issues. Do you have any additional thoughts along these lines, perhaps further analysis based on the dns-fp/id data sets? -danny _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www1.ietf.org/mailman/listinfo/dnsop
